1. Field of the Invention
The present invention relates to an automatic focus adjustment apparatus and an automatic focus adjustment method in an imaging apparatus that extracts, when an imaging target includes a predetermined object such as a human, a feature portion (face or the like) of the human, and adjusts a focus in consideration of its result.
2. Description of the Related Art
In autofocus (AF) control for automatically adjusting a focus of a video camera, the following television (TV)-AF method is normally used. That is, an AF evaluation value signal indicating sharpness (contrast state) of a video signal generated by photoelectrically converting light from an object by using an image sensor is generated, and a position of a focus lens where the AF evaluation value signal is largest is searched for.
However, in the imaging apparatus that has a face detection function, when a human is image-captured, a contrast of a human face is generally low, and therefore it is difficult to acquire a sufficient AF evaluation value. Consequently, when a distance from the object changes, identification of a focusing direction is difficult by a change of the AF evaluation value. This disables a lens position from following the object, thus causing blurring of an object image.
To solve such a problem, there has been offered a method for comparing detected face sizes with each other, determining that the object has moved in a near or far direction when the face size has increased or decreased, and driving the focus lens to follow the object (as discussed in Japanese Patent Applications Laid-Open Nos. 2009-31760 and 2008-276214).
However, in the auto-focusing the face detection, no consideration is given to an imaging parameter such as a depth of field or a state of the object, and thus the following problem may occur. That is, because of difference in size of the detected individual faces or depending on setting of a parameter during imaging, a direction of the distance change of the object may be erroneously determined, or unnecessary movement determination may be made even when the object is not blurred, thus causing reduction of autofocus accuracy.